Dan TM's Blog
Views
3545
Comments
0
 

If you thought I'd gone absolutely frigging insane last week, you ain't seen nothing yet.  I have channeled the math geek that has been lurking beneath the surface of this Theatre major/English minor, suppressed for years, to solve one of the many problems with the Natural Power Rankings.

And this is all your fault, Jstrscourt.  You made me think about how going purely by margin of victory as the second factor isn't quite right.  So I developed a formula to get a better measure of how strong a win was.  Trust me, you don't want to know how ridiculous this thing is, or how much time I spent on it.  But it works, and it's consistent.  The current scores range from .37 (Dallas over Buffalo, 25-24) to 27 (Pittsburgh over Seattle, 21-0).  The average is 5.67, which is something like Cleveland over Baltimore, 27-13.  The median, in this case a better measure of a normal score, is 3.92, the Giants' win over the Jets, 35-24. 

For first-time readers, here's how the Natural Power Rankings work.  I ignore everything I know about how good a team is supposed to be.  I look at who they've beaten and by how much, and rank the teams accordingly.  In other words:

1. If A beat B, A is higher than B.

2. If A beat C by more points than B beat C, A is higher than B.

3. If A beat C by the same margin as B beat D, and C is ranked higher than D, A is ranked higher than B.

4. If rank still cannot be determined, margin of victory is weighed against quality of opponent (by Natural Power Ranking), and whichever is more significant determines the higher ranked team. 

As a new modification last week, I only considered the past three weeks, because it got too messy otherwise.  This week, if I only consider the past three weeks it's actually messier than if I consider four, due to all the byes.  So I'm going to have to stick with four for a while, I imagine.

The problem stems from the NFC North and Washington.  Green Bay beat Washington, who beat Detroit, who beat Chicago, who beat Green Bay.  Also, Green Bay beat Minnesota, who beat Chicago, who (of course) beat Green Bay.  And none of them played anywhere else, so there's absolutely no logical way to fit them in with the rest of the league. 

But adding another week solved that problem.  And the way I've resolved such circular tangles in the past is: whoever beat the other teams in the circle by the most points gets ranked first.  Only now, instead of points, I'm using the new formula, which ranks them in this order: Washington, Chicago, Green Bay (barely behind the Bears), Minnesota, Detroit.  And now that they're in order, they cannot be split up - they are an inseparable chunk of nearly-tied teams.

Anyway, enough talk, here are the rankings, with their record over the past four weeks in parentheses:

  • 1) NY Giants (4-0)
  • 2) New England (4-0)
  • 3) Philadelphia (2-1)
  • 4) Dallas (3-1)
  • 5) Buffalo (1-2)
  • 6) Jacksonville (3-0)
  • 7) Kansas City (2-1)
  • 8) Washington (1-2)
  • 9) Chicago (1-3)
  • 10) Green Bay (3-1)
  • 11) Minnesota (1-2)
  • 12) Detroit (1-2)
  • 13) San Diego (2-2)
  • 14) Indianapolis (3-0)
  • 15) Tampa Bay (3-1)
  • 16) Tennessee (2-1)
  • 17) Oakland (2-1)
  • 18) Cleveland (2-2)
  • 19) Carolina (3-1)
  • 20) New Orleans (1-2)
  • 21) Atlanta (1-2)
  • 22) Denver (0-3)
  • 23) Houston (1-3)
  • 24) NY Jets (1-3)*
  • 25) Miami (0-4)
  • 26) Baltimore (3-1)
  • 27) Arizona (2-2)
  • 28) Pittsburgh (2-1)
  • 29) Seattle (2-2)
  • 30) Cincinnati (0-3)
  • 31) San Francisco (0-3)
  • 32) St. Louis (0-4)

* The Jets are a problem.  They played Philadelphia much closer than Detroit did, so they should be up there right behind the Bills.  They also played the Giants closer than the Eagles did, and only lost to the Bills by 3.  But their win over Miami was ever so slightly less impressive than Houston's (purely score-wise).  So they should be down there.  I'm going to establish a new rule: where there is a conflict like this, the team will be placed next to the team they finished closest to.  In this case, that means they're right above the Dolphins.

According to these rankings, here's who will win this weekend.  It's worth noting that my Natural Power Rankings went 9-4, correctly predicting the New Orleans, Tampa Bay, Carolina, and Kansas City upsets.  I went 7-6, missing three of those, but also getting the Jacksonville win right, which the NPR missed.

Buffalo beats Baltimore

Washington beats Arizona

New Orleans beats Atlanta

NY Giants beat San Francisco

New England beats Miami

Tennessee beats Houston

Detroit beats Tampa Bay

Kansas City beats Oakland

NY Jets beat Cincinnati

Seattle beats St. Louis

Philadelphia beats Chicago

Dallas beats Minnesota

Denver beats Pittsburgh

Jacksonville beats Indianapolis

Comment

Remember to keep your posts clean. Profanity will get filtered, and offensive comments will be removed.


Start Your Own Blog

Start Now

Truth & Rumors

MOST POPULAR

  1. 1
    Clippers, Warriors exchange barbs
    Views
    477
    Comments
    237
  2. 2
    Tuukka Rask takes blame for Bruins' Game 1 loss
    Views
    2485
    Comments
    231
  3. 3
    Smush Parker allegedly punches high schooler
    Views
    2099
    Comments
    140
  4. 4
    Time to penalize NHL's perennial losers?
    Views
    379
    Comments
    129
  5. 5
    Quarterback freefalling down draft boards
    Views
    7202
    Comments
    90

SI.com

SI Photos